Home » Posts » The 3 Disciplines of MI #1 : Visual Signals

The 3 Disciplines of MI #1 : Visual Signals

2510_albert_einstein
Dr Albert Einstein

Einstein is thought to have said: “If I can’t picture it, I can’t understand it”.

Whether he really said it, using more visual management information is a fast, practical step to deeper understanding and productive leadership conversations.

Our claim is that making management information more visual not only improves decision-making but shifts the agenda towards improving strategic value.

This is the first of the 3 Disciplines of Management Information : Visual Signals.

The 3 Disciplines of Management Information

These are:

Discipline #1 Visual Signals.
Discipline #2 Meaningful Measures   
Discipline #3 Strategic Alignment

What seems to surprise people most about The 3 Disciplines of MI is that they have nothing to do with Business Intelligence (BI) tools, data warehouses or technology at all. Even the types of business data are largely irrelevant. These disciplines are simple, practical and fast. Apply them and you’ll start a journey to Management Information* mastery.

*By Management Information (MI), we mean a reduction in uncertainty so that business leaders can focus on improving the things which matter and provide evidence of it.

Discipline #1 Visual Signals

In this, the first of 3 articles, we make the case for Discipline #1 : Visual Signals. We show how business leaders can make changes to their Management Information so they can see true performance signals and make evidence-based decisions.

  1. The visual content is focused on the strategic improvements and decisions which need to be made by an executive leadership audience.
  2. Management Information is presented in a consistent visual way using chart types which detect signals and exceptions rather than obscuring them.

An important factor is the decision setting and the feedback evidence needed in each context. We’re focusing on the situation where leadership teams are reporting upwards towards a board or exec committee which may also be reporting to other stakeholders.

These leadership teams are trying to make a difference in something which matters and need credible feedback about whether their investments are working. What these leaders need is management information which tells them that results and outcomes are truly getting better.

What does your Management Reporting Look Like?

Over the years I’ve run both smaller businesses and advised larger ones. I wonder if your experience of management reporting is anything like mine?

What actually gets looked at and talked about is mostly non-visual, or the visual elements are home-grown and spread across several documents. The message is fragmented and buried in text and tables of numbers to which people apply their own interpretations. Sometimes there’s a little bit of C.Y.A. in there too. There is literally no overall picture with a line of sight to the strategic objectives even if we know what they are.

The conversations and any metrics usually concern events, activity, milestones, actions, targets and compliance. These are no doubt important in some settings but are not strong evidence of progress towards strategic goals. It’s rare that the drivers of strategic performance – especially the intangible ones – are explicitly tracked.

Sources of data are fragmented too. Finance, Sales, Marketing, HR, IT have point-solutions which may be great at the functional jobs they do. But their reports all look different and don’t allow comparative oversight across business areas which are in fact intimately related.

A New Visibility

Signal Snap Viz
A Simple A4 Visual Management Pack

Imagine instead a single view of performance which forms the focal point of the management reporting and discussion. Now everything that matters to this leadership team – performance, decisions, risk – is visible in one place. With a single sheet of paper (or screen) it is now possible to compare relative performance across business areas.  A consistent use of visual design reduces the interpretive effort and produces rapid insight.

Only the important summary and exception information is shown and there’s a high level of integrity in the signals and their interpretation. There is now clarity in the message, the improvement story and evidence-based reasoning. Initiatives can now be seen to be driving strategic shifts towards goal achievement and business alignment. Shifting these performance signals now becomes the focus of the conversations and the decision-making.

Often at this point, someone will point out that they already have an Executive Dashboard or Scorecard. Of course this is great because it shows recognition of the need for comparative oversight of performance. As you’ll see these types of ‘traffic light’ reports can be improved dramatically using better visual signals or made more meaningful and strategic using the other two disciplines.

Lessons in Visual Perception

Human evolution means that we are highly tuned to seeking visual patterns and cues. Most Management Information in the form of text and tables doesn’t exploit this instinct and the use of poor visual design can distort it. If we want to interpret patterns reliably and extract meaning from them we have to pay attention to visual design.

peace-on-earth
We do fancy visualisation too.

We need good charts in management reports. Why? Because good charts show patterns which lead to rapid understanding and an increased state of knowledge. In contrast, bad charts – or none at all – expose us to more uncertainty and greater chance of making bad decisions.

There’s still some misunderstanding about how to use data visualisation in different contexts. We’re seduced by attention-grabbing marketing infographics, glossy mobile apps and interactive data visualisations (we do these at Measureology too). These all have their own purpose but they don’t show leadership teams the most important performance signals.

Like any work, we need to select the right visual tools for the job and this is where discipline is needed. Its very easy to make poor choices when confronted with the visual equivalent of self-service fast-food. Its perhaps appropriate that pie and doughnut charts are the most common examples of this visual bloat.

There’s an excellent body of research to draw on but software developers and users aren’t often aware of the good visualisation practice or how to apply it. There is increasing recognition in the data visualisation community thanks to the work of Edward Tufte, Stephen Few and Colin Ware.

We’ve applied this research at Measureology and distilled in into our first discipline of what makes good visual Management Information.

3 Gotchas with Management Reporting

There are 3 routine mistakes in management reporting which obscure the insights needed for decision-making. Fortunately, each of these can be resolved by the use of good visual practices.

  1. Not showing change over time. Strategic improvement is change. Change is something which happens over time. Improvement can’t be
    bsc
    The 3 Gotchas get Balanced Scorecards too

    shown using a single value this month, an accumulated year to date or an infrequent quarterly or annual snapshot.

  2. Using invalid comparisons. Given 2 numbers, comparing them may seem like an obvious thing to do. This month vs. last month or the same month last year, ranking of one thing over another, comparison to a target, variance from a budget. Not only are these statistically dodgy but also have behavioural dark sides.
  3. Neglecting the shape of data. Most common here is the flawed use of mean averages which mask the distribution of values. This creates the illusion of certainty about a single number where the underlying pattern may be very different.

These are often accountancy legacies being applied in an uncertain world which is better described with statistics. Money is tangible and countable in retrospect but intangible drivers in the real world exhibit natural random variability. Applying these practices to non-financial data can produce signals which aren’t really there and miss signals which really are.

3 Charts Which Fix Them

It sounds trivial to say that these problems can be fixed using better charts but to an extent they really can. Perhaps even more surprising is that we only need 3 types of chart to do it.

  1. XmR Chart. This is essentially a line chart which shows how a performance measure is changing over time. By adding statistical process
    Bullet Chart
    A Bullet Chart

    control limits we can detect true changes in a measure with high confidence and also predict its most likely future values.

  2. Bullet Chart. This is essentially a bar chart but set against a graduated background of thresholds and a target indicator. The bullet chart is
    most robust when the measure, thresholds and targets have a statistical origin.
  3. Distribution Chart. Commonly a histogram showing frequency or cumulative distribution but could also be distilled into a box plot or range chart for space-saving comparisons.

Building the first prototype report based on existing data is a very quick way to get started with these visual practices. This establishes good chart design standards and at the same time nudges forward a more valid statistical interpretation.

Telling a Richer Story

Signal Snap Viz 4
A3 Visual Management Pack for a CIO

Science tells us that we can compare information more easily within a 5 degree eye span. In other words we should use a single page or screen to group relatable visual elements together.

The layout of charts on the page or screen also tells a story about priority and the flow of causal influence from activities to strategies to
objectives.

Our 3 fundamental chart types can be adapted and combined in a very dense display to communicate these rich messages.

XmR charts can be stripped back to sparklines, bullet charts become horizontal bars, histograms are pareto charts, box-plots become percentile range plots. Building a visual reporting pack around these 3 consistent chart types makes them directly accessible and comparable.

Spot Lights and Traffic Lights

Another part of this visual discipline is in the appropriate use of colour in management reports to avoid us being blinded to the true signals.

Cognitive science tells us that we should be careful where we point the attentional spotlight to avoid our faculties being consumed by noise. We should use neutral colour schemes and only highlight important things with a pop-out action colour. It is also very easy to create faulty brain-bending associations using colour if inconsistently applied.sample-kpi-scorecard

The colours that we might expect to see on scorecards and dashboards are red amber and green. Aside from the problems these can cause for the colour blind, the use of traffic lights masks the patterns which tell us what performance is really doing. The use of rolled-up indicators, thresholds and trend arrows don’t usually have the necessary statistical validity for making good judgements.

Visual Decisions for Strategic Improvement

So can the discipline of visual management information improve decision-making? We believe it can and here are the 10 reasons:

  1. Seeing true performance signals means that decisions are made based on statistical evidence rather than intuition or ideology.
  2. Having oversight of the most important performance areas in one place reduces the chance of missing something.
  3. Evidence that performance is improving shows that the efforts of the leadership team are aligned with strategic value.
  4. A focus on how to shift performance results in the charts pushes the narrative and discussions about activities and events into a supporting role.
  5. Showing the natural variability of measures changes the perception of causality so that claims of ‘success’ or ‘failure’ are verified.
  6. Seeing variability in the system reveals that beliefs contain uncertainty and leads to more robust statistical, probabilistic thinking.
  7. Seeing the true gaps between the capability of a system and the specification – such as targets or SLAs – prioritises decisions about improvement investments.
  8. Seeing a visible, statistical improvement in something becomes the focus of the attention rather than the judgement about arbitrary target achievement.
  9. Questions can be asked about the validity of target-setting and how these might distort behaviour in the system.
  10. Start to shift the culture away from ‘measurement as a threat’ to ‘measurement as improvement feedback’.

More than anything, visual management information shifts the conversation towards those decisions which produce the most strategic improvement. This same visual information then provides direct, trusted feedback about whether those decisions worked or not.

Getting Started

Are you persuaded that seeing Visual Signals in your Management Information could yield some of these 10 benefits for you? If so, you can start with an easy experiment: Translate your existing management reports into a visual prototype.

We usually build and refine these visual prototypes in Excel first. This can then be adapted to an online dashboard which your audience can consume with their preferred format and devices.

A Visual Management Pack is a quick, practical exercise – one of our Visible Sprints – and only takes a couple of weeks. You could even have something ready for your next leadership meeting.

If you want to set up a discovery call to explore the idea of a Visual Management Pack then get in touch.

Get in touch